β-Catenin regulates endocardial cushion growth by suppressing p21

Abstract

Endocardial cushion formation is essential for heart valve development and heart chamber separation. Abnormal endocardial cushion formation often causes congenital heart defects. β-Catenin is known to be essential for endocardial cushion formation; however, the underlying cellular and molecular mechanisms remain incompletely understood. Here, we show that endothelial-specific deletion of β-catenin in mice resulted in formation of hypoplastic endocardial cushions due to reduced cell proliferation and impaired cell migration. By using a β-catenin ^DM allele in which the transcriptional function of β-catenin is selectively disrupted, we further reveal that β-catenin regulated cell proliferation and migration through its transcriptional and non-transcriptional function, respectively. At the molecular level, loss of β-catenin resulted in increased expression of cell cycle inhibitor p21 in cushion endocardial and mesenchymal cells in vivo. In vitro rescue experiments with HUVECs and pig aortic valve interstitial cells confirmed that β-catenin promoted cell proliferation by suppressing p21. In addition, one savvy negative observation is that β-catenin was dispensable for endocardial-to-mesenchymal fate change. Taken together, our findings demonstrate that β-catenin is essential for cell proliferation and migration but dispensable for endocardial cells to gain mesenchymal fate during endocardial cushion formation. Mechanistically, β-catenin promotes cell proliferation by suppressing p21. These findings inform the potential role of β-catenin in the etiology of congenital heart defects.

Figure 1.. Deletion of β-catenin in endocardium disrupts endocardial cushion formation.

(A) Immunostaining on heart sections from E9.5 control (β-cat^f/f or β-cat^f/+) and β-cat^eKO (β-cat^f/f:Tie2^Cre) embryos with β-catenin antibody (red) and isolectin B4 (IB4) was co-stained to mark the endocardium (green). (B) H&E staining of E9.5 embryo sections. (C) Quantification of mesenchymal cell number within atrioventricular cushions. n = 5/group. (D) EdU labels proliferating cells and immunostaining of PECAM1 marks the endocardium. (E) Quantification of proliferating endocardial (Endo), mesenchymal (Mes), and myocardial (Myo) cells. n = 3/group. (B, F) Quantitative analysis of cell distribution showed in (B) shows the percentage of mesenchymal cells near (sub-Endo) or far away from (sub-Myo) the endocardium. n = 4/group. # < 0.01; Δ < 0.001 by unpaired t test.

Figure S1.. Deletion of β-catenin delays embryogenesis.

Representative images of fresh isolated embryos show the gross view of control (β-cat^f/f or β-cat^f/+) and β-cat^eKO (β-cat^f/f:Tie2^Cre) embryos at E9.5 and E10.5. More than five embryos were analyzed for each genotype.

Figure S2.. Deletion of β-catenin does not affect OFT cushion formation.

Representative images of HE staining show the morphology of outflow tract cushions (indicated by stars) in E9.5 control and β-cat^eKO (β-cat^f/f:Tie2^Cre) embryos.

Figure 2.. β-catenin loss impedes mesenchymal cell migration.

(A) AVC tissues from E9.5 embryos were cultured in collagen gel for 48 h. The explants were stained with phalloidin for F-actin. (B) Bar graph shows the number of cells that migrated into the gel (B). (C, D, E, F) and cell tracking analysis of time-lapse images from the AVC explants shows velocity (D), mean square displacement (E), and directionality (F) of the migratory cells using DiPer. Unpaired t test was used for the statistical calculation. Δ < 0.001.

Figure 3.. β-catenin loss disrupts filopodia formation.

(A) Immunostaining of phospho-β-catenin (red) and IB4 on tissue sections of E9.5 embryos. IB4 staining (green) shows the transforming endocardial cells within the AVC region of control embryos having well-formed membrane protrusions with enriched phospho-β-catenin (arrowhead), whereas the β-cat^eKO embryos lack these features. (B) A carton depicts the expression pattern of phospho-β-catenin. (C) Representative images from whole mount staining of E9.5 embryos with αSMA (red) and IB4 (green) antibodies show migrating mesenchymal cells within AVC region of control embryos possessing prominent filopodia (arrowhead), whereas these structures are absent in the β-cat^eKO embryos. (D) Representative images from phalloidin staining for F-actin show migrating mesenchymal cells in control embryos having distinct microspikes (arrowhead), whereas these structures are not present in the β-cat^eKO embryos.

Figure 4.. β-catenin regulates cell proliferation and migration through its transcriptional and non-transcriptional function, respectively.

(A) H&E staining shows AVC morphology of E9.5 control (β-cat^f/f or β-cat^f/+), β-cat^eKO/eKO (β-cat^f/f:Tie2^Cre), and β-cat^eKO/DM (β-cat^f/DM:Tie2^Cre) embryos. β-catenin^DM allele was used to separate the transcriptional and non-transcriptional functions of β-catenin. (B) Quantification of mesenchymal cell number within atrioventricular cushions. n = 4/group. (A, C) Quantitative analysis of cell distribution showed in (A) shows the percentage of mesenchymal cells near (sub-Endo) or away from (sub-Myo) the endocardium. n = 4/group. (D) Ki67 antibody staining labels proliferating cells (green), and IB4 immunostaining marks the endocardium (red). (E) Quantification of proliferating rate of endocardial (Endo) and mesenchymal (Mes) cell. n = 5/group. One-way ANOVA was used for the statistical calculation. # < 0.01; Δ < 0.001. (F) Representative images of whole-mount staining of E9.5 embryos with αSMA and IB4 antibodies show filopodia in the migrating mesenchymal cells within AVC region of the control and β-cat^eKO/DM embryos (arrowhead), whereas they are not present in β-cat^eKO/eKO embryos.

Figure S3.. Molecular characterization of the cushion phenotype in β-catenin mutant embryos.

AVC tissues were collected from E9.5 control, β-cat^eKO/eKO, and β-cat^eKO/DM embryos. qRT-PCR was performed to determine the mRNA level of candidate genes involved in EndoMT and endocardial cushion formation. The expression of Gapdh was used as an internal control. n = 3/group. Unpaired t test was used for the statistical calculation. * < 0.05; # < 0.01; Δ < 0.001.

Figure S4.. β-catenin loss does not affect NOTCH and BMP signaling.

Tissue sections of E9.5 embryos were subjected to immunostaining with antibodies for N1ICD (active NOTCH1 in nuclei) and phosph-SMAD1/5/9 (downstream effector of BMP signaling). The representative images show the expression of indicated markers within AVC region. The arrowheads indicate the positive signal within cushion endocardial cells for indicated markers. At least four embryos were analyzed for each staining. a, atrium; v, ventricle.

Figure 5.. β-catenin negatively regulates p21 expression.

(A) AVC tissue was microdissected from E9.5 control, β-cat^eKO/eKO, and β-cat^eKO/DM embryos and subjected to qRT-PCR analysis of p21 mRNA level. The expression of p21 mRNA was normalized to that of Gapdh. n = 3/group. (B) RNAscope analysis shows p21 mRNA expression within AVC region of E9.5 hearts. The arrowheads indicate the expression of p21 mRNA. (C, D) Immunostaining for p21 (red). The percentage of p21-expressing cells (arrowhead) was quantified (D). EC, endocardial cell; Mes, mesenchymal cell; Myo, myocardial cell. n = 4/group. (E) HUVEC was treated with 1 μM XAV939 or DMSO for 24 h. XAV939 is an inhibitor for canonical WNT/β-catenin signaling. Immunostaining shows p21 protein expression. The percentage of p21-positive cells was quantified and presented in the bar chart on right. n = 3/group. (F) Pig aortic valve interstitial cells were treated with 2.5 μM XAV939 or DMSO for 24 h. The cells were subjected to qRT-PCR analysis of p21 mRNA expression. n = 4/group. Unpaired t test and one-way ANOVA were used for the statistical calculation among two and three groups, respectively. # < 0.01; Δ < 0.001.

Figure 6.. β-catenin promotes cell proliferation by suppressing p21.

(A, B) HUVEC was transfected with control or gene-specific siRNA. qRT-PCR detected the mRNA expression of β-catenin and p21. n = 3/group. (C) HUVEC was transfected with indicated siRNA. EdU was used to label the proliferating cells. The percentage of EdU-positive cell was quantified. n = 3/group. (D) PAVIC were treated with XAV939 (2.5 μM) and UC2288 (2.5 μM). UC2288 is an inhibitor for p21. Cell proliferation was detected by EdU labeling and quantified. n = 4/group. Unpaired t test and one-way ANOVA were used for the statistical calculation among two and three groups, respectively. * < 0.05; # < 0.01; Δ < 0.001.

Figure S5.. β-catenin negatively regulates the protein level of VE-cadherin.

(A) Immunostaining shows the expression of VE-cadherin protein (red) within the hearts of E9.5 embryos. The upper and lower panel shows the images from the AVC cushion and ventricular region on the same section, respectively. (B) RNAscope analysis shows the mRNA expression of VE-cad (red) within AVC cushion region of E9.5 hearts. (C) qRT-PCR analysis of the mRNA level of VE-cad in the AVC tissues from E9.5 embryos. The expression of VE-cad was normalized to that of Gapdh. n = 3/group. One-way ANOVA was used for statistical calculation.

Figure 7.. β-catenin is dispensable for endocardial-to-mesenchymal fate change.

(A, B) Tissue sections of E9.5 embryos were subjected to immunostaining with antibodies for mesenchymal markers (αSMA, VIMENTIN, PDGFRβ) (A) or transcriptional factors (SNAIL, SLUG, and SOX9) involved in EndoMT process (B). The representative images show the expression of indicated markers within AVC region. The arrowheads indicate cushion endocardial cells. At least four embryos were analyzed for each staining. a, atrium; v, ventricle.

Figure 8.. β-catenin regulates cell proliferation and migration essential for endocardial cushion formation.

(A) Schematic shows the distinct cushion defects in β-cat^eKO/eKO and β-cat^eKO/DM embryos. At E9.5, cushion endocardial cells within AVC region of control embryos undergo EndoMT, and then the transformed mesenchymal cells delaminate from the endocardial sheet, invade into the cardiac jelly, and proliferate, generating endocardial cushions which serve as the valve primordia and eventually give rise to atrioventricular valves. In contrast, loss of β-catenin inhibits proliferation of endocardial and mesenchymal cells, leading to formation of hypocellular endocardial cushions in both β-cat^eKO/eKO and β-cat^eKO/DM embryos. The underlying mechanism involves up-regulation of cell cycle inhibitor p21. In addition, complete loss of β-catenin in β-cat^eKO/eKO embryos impairs filopodia formation and impedes mesenchymal cell migration away from the endocardial sheet. However, selective disruption of the transcriptional function of β-catenin β-cat^eKO/DM embryos does not affect cell migration. (B) β-catenin suppresses p21 to promote cell proliferation. On the other hand, β-catenin is essential for filopodia formation and cell migration.